Skills Registry

VerifiedSafe

Dynamic skill discovery and matching for any task context.

Sby Skills Guide Bot
ProductivityIntermediate
7/26/2026
Claude Code
#skill-discovery#skill-matching#keyword-matching#agent-orchestration#meta-skill

Recommended for

Automated summary

Analyzes the current task, extracts keywords, and matches them against an internal skill registry to recommend which skills to apply.

Strengths

  • Structured context extraction across four axes: action, domain, technology, and problem type
  • Ranks results by priority and enforcement level (required before suggested)
  • Explicit "no skills matched" fallback that asks for the domain or stack instead of forcing a match
  • Limited to Read, Grep, and Glob, so recommendations stay read-only

Limitations

  • Only knows the skills listed in its own registry table, which must be maintained by hand
  • Matching is keyword-based, so phrasing without technical vocabulary may trigger nothing
  • The end of the SKILL.md is truncated, so some registry entries may be missing
When to use it

Use it at the start of a task, or when an agent or command needs to know which specialized skills to pull in.

When not to use it

Skip it if the skill registry is not kept up to date, or if you already know exactly which skill to apply.

Security analysis

Safe
Quality score80/100

This is a pure documentation/routing skill: it defines a keyword-matching protocol over a static table of local skill paths and uses only read-only file tools. Nothing in it executes commands, touches the network, or handles secrets. The only non-trivial behavior is instructing the agent to load and apply other skill files, which is a legitimate registry pattern but does forward trust to whatever those files contain.

Findings
  • •Declared tools are read-only (Read, Grep, Glob); no bash, network, or write capability requested.
  • •The skill directs the agent to recommend and 'apply' other skills by hardcoded path (.claude/skills/...), which can chain-load additional instruction files whose contents this file does not constrain — a mild trust-propagation surface worth noting, though the referenced paths are local repo paths.
  • •No obfuscation, no credential/secret handling, no destructive commands, no safety-disabling language.
How this analysis is made

Examples

TDD with zod
Match skills for a TDD cycle with zod validation on the payments service.
React landing page
What skills should I apply for a React landing page with animated hero sections?
Debugging a migration
Find skills for debugging a failing Prisma migration in our monorepo.

name: skills-registry description: | Dynamic skill discovery and matching for any task context.

Invoke when:

  • Starting any task to discover relevant skills
  • Need to find specialized skills for current context
  • Agent/command needs skill recommendations

Trigger phrases: "match skills", "find skills", "what skills for", "recommend skills", "which capabilities", "skill lookup", "applicable skills"

Examples:

  • "Match skills for TDD cycle with zod validation" → outputs backend-zod, backend-vitest
  • "What skills for React landing page?" → outputs frontend-master, frontend-aceternity allowed-tools: Read, Grep, Glob

Skills Registry

Dynamic skill matching system. Analyzes task context and recommends skills to apply.

Matching Protocol

Step 1: Extract Context Keywords

From current task/input, identify:

  • Action: analyze, fix, create, test, debug, review, implement, setup, deploy
  • Domain: frontend, backend, API, database, validation, UI, auth, DevOps
  • Tech: React, Next.js, Express, Prisma, tRPC, zod, Playwright, Docker
  • Problem: error, bug, issue, failure, lint, type check, debugging

Step 2: Match Against Registry

For each skill in registry:

  1. Count keyword matches: (task keywords) ∩ (skill keywords)
  2. Check if task situation matches skill's when condition
  3. Match if: ≥2 keywords OR when applies

Step 3: Rank Results

Order matched skills by:

  1. priority: high → top of list
  2. enforcement: required → above suggest
  3. More keyword matches → higher rank

Step 4: Output

When matches found:

Matched skills:

**Required:**
- Apply [Skill] skill — [reason]

**Suggested:**
- Apply [Skill] skill — [reason]

When no matches (< 2 keywords, no when applies):

No skills matched.
Context extracted: [keywords]
Clarify domain or tech stack for better matching.

Skills Registry Data

Core Skills

| Skill | Path | Keywords | When | Priority | Enforcement | |-------|------|----------|------|----------|-------------| | feature-analyzer | .claude/skills/feature-analyzer/ | feature, artifacts, spec, plan, tasks, data-model, requirements, context, ux, contracts | Loading feature context OR reading documentation artifacts OR need requirements understanding | high | required | | code-analyzer | .claude/skills/code-analyzer/ | codebase, structure, dependencies, markers, AICODE, imports, modules, architecture, existing code | Need project structure OR dependency analysis OR find existing AICODE markers OR understand architecture | medium | suggest | | git | .claude/skills/git/ | branch, commit, push, merge, git, repository, checkout, stash | Any git operation — branch creation, commits, pushes, merges | high | required | | sequential-thinking | .claude/skills/sequential-thinking/ | complex, analysis, multi-step, root cause, debugging, unclear, diagnosis, investigation | Problem requires 3+ logical steps OR unclear root cause OR multiple valid approaches OR low confidence in solution | medium | suggest | | context7 | .claude/skills/context7/ | library, documentation, package, npm, pypi, external, api reference, third-party | Need library documentation OR unfamiliar package OR library-related error OR checking correct API usage | medium | suggest | | self-commenting | .claude/skills/self-commenting/ | AICODE, marker, NOTE, TODO, FIX, annotation, cross-session, context preservation | Writing code that needs context for future sessions OR documenting non-obvious logic OR leaving markers | low | suggest |

Agent & Prompt Skills

| Skill | Path | Keywords | When | Priority | Enforcement | |-------|------|----------|------|----------|-------------| | agent-creator | .claude/skills/agent-creator/ | agent, subagent, create agent, agent template, agent design, specialized agent, domain expert | Creating new agents OR designing agent prompts OR configuring agent tools and models | medium | suggest | | self-improve | .claude/skills/self-improve/ | intent, interpret, vague request, enrich, synthesize, findings, actionable, user request | Interpreting vague user intent OR preparing context for agent generation OR synthesizing actionable insights from project context | medium | suggest | | prompt-optimizer | .claude/skills/prompt-optimizer/ | prompt, TCRO, prompt engineering, optimize prompt, improve prompt, prompt template, structure prompt | Crafting prompts for code generation OR improving prompts not getting results OR structuring multi-step workflows OR debugging pattern drift | medium | suggest |

Frontend Skills

| Skill | Path | Keywords | When | Priority | Enforcement | |-------|------|----------|------|----------|-------------| | frontend-master | .claude/skills/frontend-master/ | frontend, Next.js, React, Tailwind, UI, components, styling, layout | Starting frontend task OR need decision framework for UI components, animations, assets, or theming | high | suggest | | frontend-shadcn | .claude/skills/frontend-shadcn/ | shadcn, button, input, dialog, form, table, modal, dropdown, Radix, components | Need standard UI components (buttons, inputs, dialogs, forms, tables) OR building React/Next.js UI | low | suggest | | frontend-aceternity | .claude/skills/frontend-aceternity/ | aceternity, spotlight, aurora, 3D card, hero effects, dramatic, wow factor, landing page | Need dramatic hero effects, spotlights, aurora backgrounds, 3D hover cards, or text reveal animations | low | suggest | | frontend-magic-ui | .claude/skills/frontend-magic-ui/ | magic ui, number ticker, marquee, bento grid, mockup, Safari, iPhone, shimmer button, SaaS | Need SaaS landing components like number tickers, logo marquees, bento grids, or device mockups | low | suggest | | frontend-lottie | .claude/skills/frontend-lottie/ | lottie, animation, loader, spinner, success, error, empty state, decorative | Need simple play/loop animations like loading spinners, success checkmarks, or empty state illustrations | low | suggest | | frontend-rive | .claude/skills/frontend-rive/ | rive, interactive animation, state machine, hover, click, data-driven, toggle, checkbox | Need animations that react to user input (hover, click, drag) OR have multiple states/transitions OR respond to data values | low | suggest | | frontend-color-system | .claude/skills/frontend-color-system/ | color, palette, theme, brand color, dark mode, WCAG, contrast, accessibility | Setting up project colors OR creating dark mode OR checking WCAG contrast OR generating theme from brand color | low | suggest | | frontend-google-fonts | .claude/skills/frontend-google-fonts/ | fonts, typography, Google Fonts, font pairing, Inter, Plus Jakarta, headings, body text | Setting up project fonts OR need font pairing recommendations OR optimizing font loading | low | suggest | | frontend-iconify | .claude/skills/frontend-iconify/ | icon, iconify, lucide, heroicons, SVG, icon search, icon set | Need to find icons by concept OR integrate icon library OR download SVG icons | low | suggest | | frontend-image-generation | .claude/skills/frontend-image-generation/ | avatar, placeholder, photo, illustration, DiceBear, Unsplash, unDraw, image assets | Need avatars, placeholder photos, or illustrations — use FREE resources first (DiceBear, Unsplash, unDraw) | low | suggest | | frontend-debug-linting | .claude/skills/frontend-debug-linting/ | lint, ESLint, TypeScript, type check, Prettier, format, debug, console errors | After writing React/Next.js code OR before delivery OR debugging frontend issues | medium | required | | frontend-playwright | .claude/skills/frontend-playwright/ | playwright, screenshot, browser, visual QA, responsive, viewport, console errors, UI testing | Before delivering UI changes OR need visual verification OR responsive testing OR debugging hydration issues | low | required | | frontend-design-review | .claude/skills/frontend-design-review/ | design review, UI review, UX, visual polish, accessibility, WCAG, responsive design, component standards | Building new interfaces OR reviewing visual changes OR validating design quality OR accessibility compliance | medium | suggest |

Backend Skills

| Skill | Path | Keywords | When | Priority | Enforcement | |-------|------|----------|------|----------|-------------| | backend-master | .claude/skills/backend-master/ | backend, API, server, TypeScript backend, Express, Node.js, microservices | Starting backend task OR need decision framework for APIs, auth, database, validation, logging, testing, or deployment | high | suggest | | backend-trpc | .claude/skills/backend-trpc/ | tRPC, type-safe API, procedure, router, query, mutation, middleware, end-to-end types | Building full-stack TypeScript API OR need type-safe API without GraphQL OR building internal APIs | low | suggest | | backend-trpc-openapi | .claude/skills/backend-trpc-openapi/ | OpenAPI, REST, Swagger, API documentation, third-party integration, external clients | Need REST endpoints from tRPC OR Swagger documentation OR third-party integrations need REST | low | suggest | | backend-zod | .claude/skills/backend-zod/ | zod, validation, schema, input validation, type inference, parse, safeParse, coerce | Validating API inputs OR form data OR environment variables OR any data crossing trust boundaries | low | suggest | | backend-prisma | .claude/skills/backend-prisma/ | prisma, database, ORM, schema, migration, PostgreSQL, MySQL, SQLite | Need database access with TypeScript types OR running migrations OR querying relational data | low | suggest | | backend-auth-js | .claude/skills/backend-auth-js/ | auth.js, NextAuth, OAuth, GitHub login, Google login, session, JWT, credentials | Building Next.js app with OAuth OR credentials login OR session management | low | suggest | | backend-passport-js | .claude/skills/backend-passport-js/ | passport, Express auth, JWT authentication, strategy, bearer token, login, register | Building Express API with JWT auth OR need custom auth strategies OR non-Next.js authentication | low | suggest | | backend-pino | .claude/skills/backend-pino/ | pino, logging, structured logs, JSON logs, request logging, child logger, redact, observability | Need structured logging for production OR request tracing OR sensitive data redaction | medium | suggest | | backend-vitest | .claude/skills/backend-vitest/ | vitest, test, unit test, mock, describe, expect, coverage, TDD | Writing unit tests OR testing tRPC procedures OR testing Zod schemas OR need fast test runner | low | suggest |

DevOps Skills

| Skill | Path | Keywords | When | Priority | Enforcement | |-------|------|----------|------|----------|-------------| | docker-node | .claude/skills/docker-node/ | docker, container, Dockerfile, docker-compose, deploy, CI/CD, multi-stage, production | Deploying Node.js application OR need consistent dev environment OR setting up CI/CD OR containerizing API | medium | suggest |


Examples

Example 1: TDD with Validation

Input: "Implementing user validation with zod schemas, need unit tests"

Output:

Matched skills:

**Suggested:**
- Apply backend-zod skill — validation, zod matched
- Apply backend-vitest skill — unit test, TDD matched

Example 2: ESLint Errors (Required Enforcement)

Input: "ESLint errors after adding new React component, type check failing"

Output:

Matched skills:

**Required:**
- Apply frontend-debug-linting skill — ESLint, type check matched (enforcement: required)

**Suggested:**
- Apply frontend-master skill — React context

Example 3: Complex Problem (Sequential Thinking)

Input: "Root cause analysis of auth failure, unclear why JWT validation fails"

Output:

Matched skills:

**Suggested:**
- Apply sequential-thinking skill — root cause, unclear, analysis matched
- Apply backend-auth-js skill — auth, JWT matched
- Apply backend-passport-js skill — JWT authentication matched

From the file to a first result

The download contains exactly the content displayed on this page. This copy may differ from the latest version in the source repository.

View the source and supporting files

This download contains SKILL.md only. Any scripts, references, templates and dependencies are not included. Get the required files from the source, preserving their directory structure, before continuing.

1. Download the file

Download SKILL.md

From the root of a test project, save the file as .skill-downloads/skills-registry-ms1aw44l/SKILL.md, or use the command below. This folder is for inspecting the skill before installation.

Command for macOS/Linux or WSL: requires a POSIX shell and curl. On Windows without WSL, use the download button and create the folder manually.

mkdir -p '.skill-downloads/skills-registry-ms1aw44l' && curl --fail --show-error --location 'https://skills-guides.com/api/skills/skills-registry-ms1aw44l/download' --output '.skill-downloads/skills-registry-ms1aw44l/SKILL.md.part' && mv '.skill-downloads/skills-registry-ms1aw44l/SKILL.md.part' '.skill-downloads/skills-registry-ms1aw44l/SKILL.md'

2. Check prerequisites

Read SKILL.md before using it. If it references scripts or external services, inspect them and prepare the required access in your test environment.

Tools declared by the author (check availability in your agent): Read, Grep, Glob

No recognized prerequisites section in this file. This does not mean it is self-contained: libraries, tools and access requirements may be mentioned elsewhere in the content or source repository.

The platforms on this page are catalog indications. The paths below follow tool documentation; they do not prove this skill has been run in those tools. No minimum tool version is confirmed for this listing.

3. Load the skill

After inspection, copy the complete folder into your project using one of the commands below, then open a new tool session. The command stops if a folder with the same name already exists: compare the files before replacing it.

Claude Code
test -f '.skill-downloads/skills-registry-ms1aw44l/SKILL.md' && test -s '.skill-downloads/skills-registry-ms1aw44l/SKILL.md' && mkdir -p '.claude/skills' && test ! -e '.claude/skills/skills-registry' && cp -R '.skill-downloads/skills-registry-ms1aw44l' '.claude/skills/skills-registry'

OpenSkills

OpenSkills must already be installed. The local install/read procedure is checked with OpenSkills 1.5.0, which requires Node.js ≥ 20.6.0. Catalog slugs are not resolved automatically.

Official OpenSkills installation
openskills install './.skill-downloads/skills-registry-ms1aw44l'

openskills read 'skills-registry-ms1aw44l'

Reading should display the installed file content and its base directory. It does not run the skill or install its dependencies.

For an agent that reads AGENTS.md, then sync the index. This command changes AGENTS.md: review the diff before using it.

openskills sync

Manual method or another tool: attach SKILL.md to a conversation, or open it in an agent that can read files. Include any required supporting resources. No universal folder is recognized by every tool.

4. Try a first task

Use a short, synthetic input: an API endpoint, a small document or an interface brief. In a new conversation, attach the file or give the agent its path, then adapt this request:

Read .skill-downloads/skills-registry-ms1aw44l/SKILL.md and apply its instructions to this example: [describe a test input]. Identify any missing files, dependencies or access. Produce the result requested by the skill without inventing data or running unreviewed scripts.

Expected result: output matching the skill instructions, or a precise list of missing prerequisites. A read confirmation alone does not validate a first use. If an external resource is unavailable, that path remains untested.

Related skills